ECON 3880: Assignment 2

1 Loading Data

df <- read.csv("C:/Users/Adam/Downloads/insurance.csv")

2 Q1:

library(ggplot2)
ggplot(df, mapping=aes(region, sex))+
  geom_count(colour="blue")

The region with the highest number of observations is the southeast. This is evident by looking at the plot and seeing the magnitude of the circles in each region, of which the southeast has 170 observations for females in the southeast, and 180 observations for males in the same region.

3 Q2:

ggplot(df, mapping=aes(x=bmi, y=charges, color=smoker))+
geom_point()+
  labs(x="Body Mass Index", y="Dollar Charged")+
  geom_smooth(method="lm", aes(group=smoker))+
  ggtitle("The link between BMI and Charges: Smokers and non-smokers")

We can conclude many things from the scatter plot above. Firstly, we can conclude that there is a positive correlation among smokers and body mass index, where as the bmi increases for the individual, the amount of insurance charges increases. Furthermore, we can conclude that there are two sections among where most of the smokers reside; being from approximately 15 to 30 bmi, and then the amount of charges increases marginally from 30 to 60bmi. In addition to this, most of the smokers, with respect to an increasing body mass index, have a drastically lower amount of insurance charges than those who are smokers, aside from the few outliers which have higher insurance charges which may be due to unobservable factors that are not in our dataset.

4 Q3:

ggplot(df, mapping=aes(x=region, fill=smoker))+
  geom_bar()+
  coord_flip()+
  theme_minimal()+
  geom_text(
    stat = "count" ,
    aes(label=stat(count)),
    hjust=1.5)+
  scale_fill_brewer(palette="Dark2")

5 Q4:

library(plotly)
plot_ly(data=df,
        x=~bmi,
        y=~charges,
        color=~region,
        type="scatter",
        mode="markers")